论文检索

Xiao Zhan, Guangzhi Sun, Jose Such, Phil Woodland

Audio Large language models (LLMs) are increasingly deployed in the real world, where they inevitably capture speech from unintended nearby bystanders, raising privacy risks that existing benchmarks and defences did not consider. We introduce SH-Bench, the first benchmark designed to evaluate selective hearing: a model’s ability to attend to an intended main speaker while refusing to process or reveal information about incidental bystander speech. SH-Bench contains 3,968 multi-speaker audio mixtures, including both real-world and synthetic scenarios, paired with 77k multiple-choice questions that probe models under general and selective operating modes. In addition, we propose Selective Efficacy (SE), a novel metric capturing both multi-speaker comprehension and bystander-privacy protection. Our evaluation of state-of-the-art open-source and proprietary LLMs reveals substantial bystander privacy leakage, with strong audio understanding failing to translate into selective protection of bystander privacy. To mitigate this gap, we also present Bystander Privacy Fine-Tuning (BPFT), a novel training pipeline that teaches models to refuse bystander-related queries without degrading main-speaker comprehension. We show that BPFT yields substantial gains, achieving an absolute 47% higher bystander accuracy under selective mode and an absolute 16% higher SE compared to Gemini 2.5 Pro, which is the best audio LLM without BPFT. Together, SH-Bench and BPFT provide the first systematic framework for measuring and improving bystander privacy in audio LLMs.

Linrui Ma, Chun Hei Lo, Xinyu Wang, Peng Lu, Xihao Yuan, Hanting Chen, Kai Han, Xinghao Chen, Chengjun Zhan, Hanlin xu 等

The quadratic computational cost of traditional attention mechanisms poses a major bottleneck to the scalability and practical deployment of large language models (LLMs), particularly in long-context scenarios. To improve efficiency, existing approaches often enforce rigid structural constraints such as local attention windows. However, these strategies typically lead to substantial performance degradation on tasks requiring precise long-range recall. In this work, we propose MATCH, a scalable and efficient framework that augments sparsified attention mechanisms with dynamically integrated in-context information through an efficient retrieval system. Empirical results show that MATCH significantly improves the performance of sparse-attention models on both synthetic and real-world natural-language tasks. These findings highlight the versatility of MATCH as a general approach for enhancing in-context retrieval capabilities while maintaining the efficiency benefits of sparse attention architectures.

Xinhe Shi, Qingcheng Zeng, Weihao Xuan, Linchao Zhu

Large language models (LLMs) exhibit uneven performance across languages. In language-specific applications, practitioners often rely on target-language corpora or cross-lingual transfer to achieve better performance. However, traditional linguistic typology, commonly used as a transfer language selection strategy in previous studies, may not align with LLM’s perception of language similarity. This work proposes **LLM-based language similarity** as a novel perspective for selecting effective fine-tuning languages. We construct a framework to quantify the similarity within each language pair through both the lenses of **language-specific performance patterns** and **cross-lingual transferability**, ultimately deriving three similarity score matrices. Moreover, we observe a counter-intuitive phenomenon: **super-additive transfer effect**, where fine-tuning on a certain language yields higher performance than fine-tuning directly on the target language. Additionally, due to the absence of an existing dataset meeting our experimental requirements, we construct and release **M4CQ-Pro** dataset, which features domain-diverse distribution of **135** tasks and content consistency across **31** languages (including over 20 medium- and low-resource languages), with 61518 manually reviewed high-quality questions per language. We evaluate our approach on representative multilingual LLMs and results show that all three LLM-based similarity measures effectively guide fine-tuning language selection, outperforming traditional linguistic similarity, with the integrated measure achieving the best results. Our approach provides not only **a novel perspective on language similarity**, but also **practical baselines for selecting fine-tuning languages**.

Xinghe Chen, Naiming Liu, Shashank Sonkar

Student mistakes in mathematics are often systematic: a learner applies a coherent but wrong procedure and repeats it across contexts. We introduce MalruleLib, a learning-science-grounded framework that translates documented misconceptions into executable procedures, drawing on 67 learning-science and mathematics education sources, and generates step-by-step traces of malrule-consistent student work. We formalize a core student-modeling problem as Malrule Reasoning Accuracy (MRA): infer a misconception from one worked mistake and predict the student’s next answer under cross-template rephrasing. Across nine language models (4B - 120B), accuracy drops from 66% on direct problem solving to 40% on cross-template misconception prediction. MalruleLib encodes 101 malrules over 498 parameterized problem templates and produces paired dual-path traces for both correct reasoning and malrule-consistent student reasoning. Because malrules are executable and templates are parameterizable, MalruleLib can generate over one million instances, enabling scalable supervision and controlled evaluation. Using MalruleLib, we observe cross-template degradations of 10 - 21%, while providing student step traces improves prediction by 3 - 15%. We release MalruleLib as infrastructure for educational AI that models student procedures across contexts, enabling diagnosis and feedback that targets the underlying misconception.

Bolei Ma, Yusuke Miyao

Do Large Language Models (LLMs) genuinely grasp the compositional semantics of events, or do they rely on surface-level probabilistic heuristics? We investigate the Imperfective Paradox, a logical phenomenon where the past progressive aspect entails event realization for activities (e.g., running \to ran) but not for accomplishments (e.g., building \nrightarrow built). We introduce ImperfectiveNLI, a diagnostic dataset designed to probe this distinction across diverse semantic classes. Evaluating state-of-the-art open-weight models, we uncover a pervasive Teleological Bias: models systematically hallucinate completion for goal-oriented events, even overriding explicit textual cancellation. Prompting interventions partially reduce this bias but trigger a calibration crisis, causing models to incorrectly reject valid entailments for atelic verbs. Representational analyses further show that while internal embeddings often distinguish progressive from simple past forms, inference decisions are dominated by strong priors about goal attainment. Taken together, our findings indicate that these current open-weight LLMs operate as predictive narrative engines rather than faithful logical reasoners, and that resolving aspectual inference requires moving beyond prompting toward structurally grounded alignment.

Jaspreet Ranjit, Hyundong Justin Cho, Claire J. Smerdon, Yoonsoo Nam, Myles Phung, Jonathan May, John R. Blosnich, Swabha Swayamdipta

The National Violent Death Reporting System (NVDRS) documents suicides in the United States. In a demanding public health data pipeline, annotators manually extract structured information from death investigation records following extensive codebooks (i.e. annotation guidelines) painstakingly developed by experts. In this work, we facilitate data-driven insights from the NVDRS data to support the development of novel suicide interventions by leveraging language models (LM) as assistants to these (a) data annotators and (b) experts. We find that LM predictions match existing data annotations about 85% of the time across 50 NVDRS variables. Where the LM disagrees with existing annotations, our expert review identifies that 38% of these instances reveal inconsistencies between narratives and structured data. Finally, we introduce a human-in-the-loop algorithm that helps experts efficiently build and refine codebooks for new variables by having them only focus on providing feedback for incorrect LM predictions. We apply our algorithm to a real-world case study, and find that about 96K narratives contain evidence of victim interactions with legal professionals, which surfaces a substantial opportunity for upstream intervention that is not captured in the original structured data. Our findings provide evidence that LMs can serve as effective assistants to public health researchers who handle sensitive data in high-stakes scenarios.

Xiutian Zhao, Björn Schuller, Berrak Sisman

Emotion is a central dimension of spoken communication, yet, we still lack a mechanistic account of how modern large audio-language models (LALMs) encode it internally. We present the first neuron-level interpretability study of emotion-sensitive neurons (ESNs) in LALMs and provide causal evidence supporting the existence of such units in Qwen2.5-Omni, Kimi-Audio, and Audio Flamingo 3. Across these three widely used open-source models, we compare frequency-, entropy-, mean-deviation-, and contrast-based neuron selectors on multiple emotion recognition benchmarks. Using inference-time interventions, we reveal a consistent emotion-specific signature: deactivating neurons selected for a given emotion disproportionately degrades recognition of that emotion while largely preserving other classes, whereas targeted steering amplifies these units to bias predictions toward the target emotion. These effects arise with modest amounts of identification data and scale systematically with intervention strength. We further observe that ESNs exhibit non-uniform layer-wise clustering with partial cross-dataset transfer. Taken together, our results offer a causal, neuron-level account of emotion decisions in LALMs and highlight targeted neuron interventions as an actionable handle for controllable affective behaviors.

Yingshan Susan Wang, Linlu Qiu, Zhaofeng Wu, Roger P. Levy, Yoon Kim

Grammaticality and likelihood are distinct notions in human language. Pretrained language models (LMs), which are probabilistic models of language fitted to maximize corpus likelihood, generate grammatically well-formed text and discriminate well between grammatical and ungrammatical sentences in tightly controlled minimal pairs. However, their string probabilities do not sharply discriminate between grammatical and ungrammatical sentences overall. But do LMs implicitly acquire a grammaticality distinction distinct from string probability? We explore this question through studying pretrained LMs’ internal representations, by training a linear probe on a dataset of grammatical and (synthetic) ungrammatical sentences obtained by applying perturbations to a naturalistic text corpus. We find that this simple grammaticality probe generalizes to human-curated grammaticality judgment benchmarks and outperforms LM probability-based grammaticality judgments. When applied to semantic plausibility benchmarks, however, in which both members of a minimal pair are grammatical and differ in only plausibility, the probe performs worse than string probability. The English-trained probe also exhibits nontrivial cross-lingual generalization, outperforming string probabilities on grammaticality benchmarks in numerous other languages. Additionally, probe scores correlate only weakly with string probabilities. These results collectively suggest that pretrained LMs acquire to some extent an implicit grammaticality distinction within their hidden layers.

Xuanwen Ding, Chengjun Pan, Zejun Li, Jiwen Zhang, Siyuan Wang, Zhongyu Wei

Evaluating multimodal large language models (MLLMs) is becoming increasingly expensive as benchmarks grow in scale and cross-modality complexity. Inspired by structuralism in cognitive psychology, we tackle this difficulty with an adaptive evaluation framework for efficient benchmarking, namely **AutoJudger**. Instead of passively scoring on a fixed test set, AutoJudger treats evaluation as an interview-like process by keeping a hypothesized ability structure of the evaluated model and actively selecting the informative questions so as to refine these ability boundaries. Specifically, AutoJudger has three core components: **ability decomposition** to organize evaluation along meaningful capability dimensions, **ability estimation** to maintain an up-to-date quantitative profile of the model competence, and **adaptive question selection** to choose the most informative questions. To operationalize this paradigm, we introduce **A^2-Judger**, a novel MLLM-based **A**gentic instantiation of **A**uto**Judger** equipped with semantic-aware retrieval and dynamic memory. Experiments on four representative multimodal benchmarks show that A^2-Judger significantly improves sample efficiency while maintaining reliable evaluation results.

Zechen Li, Baiyu Chen, Hao Xue, Flora D. Salim

Motion sensor time-series are central to Human Activity Recognition (HAR), yet conventional approaches are constrained to fixed activity sets and typically require costly parameter retraining to adapt to new behaviors. While Large Language Models (LLMs) offer promising open-set reasoning capabilities, applying them directly to numerical time-series often leads to hallucinations and weak grounding. To address this challenge, we propose ZARA (Zero-training Activity Reasoning Agents), a knowledge- and retrieval-augmented agentic framework for motion time-series reasoning in a training-free inference setting. Rather than relying on black-box projections, ZARA distills reference data into a statistically grounded textual knowledge base that transforms implicit signal patterns into verifiable natural-language priors. Guided by retrieved evidence, ZARA iteratively selects discriminative cues and performs grounded reasoning over candidate activities. Extensive experiments on eight benchmarks show that ZARA generalizes robustly to unseen subjects and across datasets, demonstrating strong transferability across heterogeneous sensor domains. These results mark a step toward trustworthy, plug-and-play motion understanding beyond dataset-specific artifacts. Our code is available at https://github.com/zechenli03/ZARA.

Xiangfeng Wang, Hangyu Guo, Yanlin Lai, Mitt Huang, Liang Zhao, Chengyuan Yao, Yinmin Zhang, Qi Han, Xiaoxiaoren, Chun Yuan 等

While model-based verifiers are essential for scaling Reinforcement Learning with Verifiable Rewards (RLVR), current outcome-centric verification paradigms primarily focus on the consistency between the final result and the ground truth, often neglecting potential errors in the derivation process. This leads to assigning positive rewards to correct answers produced from incorrect derivations. To bridge this gap, we introduce **PRIME**, a benchmark for evaluating verifiers on **PR**ocess-outcome alignment verification **I**n **M**athematics and **E**ngineering. Curated from a comprehensive collection of college-level STEM problems, **PRIME** comprises 2,530 high-difficulty samples through a consistency-based filtering pipeline. Through extensive evaluation, we find that current verifiers frequently fail to detect derivation flaws. Furthermore, we propose a process-aware RLVR training paradigm utilizing verifiers selected via **PRIME**. This approach substantially outperforms the outcome-only verification baseline, achieving absolute performance gains of **8.29%**, **9.12%**, and **7.31%** on AIME24, AIME25, and Beyond-AIME, respectively, for the Qwen3-14B-Base model. Finally, we demonstrate a strong linear correlation (R^2 > 0.92) between verifier accuracy on **PRIME** and RLVR training effectiveness, validating **PRIME** as a reliable predictor for verifier selection.

Yujan Ting, Xu Tang, Terrence Chen, Weijing Huang

Despite recent progress in context compression, we identify a fundamental memorization-utilization gap where models can compress context with near-perfect fidelity yet fail to effectively utilize these compressed representations for downstream tasks. We address this with a holistic training paradigm spanning pretraining, instruction tuning, and reinforcement learning, built upon an average pooling compression. Our key innovation uses outcome-based RL to enable implicit expansion: the model learns to adaptively unfold task-relevant details during generation, interleaving reconstruction with reasoning. We achieve near-lossless 16x context compression (≈5.3x decoder sequence-length reduction in our current implementation) across 7B and 32B models, recovering over 98% of full-context QA performance and outperforming prior methods by 11 points. Our 32B model demonstrates strong out-of-distribution and length generalization, robustly scaling to 120k-token contexts despite training on no more than 4k tokens, matching full-context performance on NIAH, LongBench v2, and multi-hop reasoning. We verify the implicit expansion behavior in experiments.

Yuanhao Ding, Meimingwei Li, Esteban Garces Arias, Matthias Aßenmacher, Christian Heumann, Chongsheng Zhang

The quality of text generated by large language models depends critically on the decoding sampling strategy. While mainstream methods such as Top-k, Top-p, and Min-p achieve a balance between diversity and accuracy through probability-space truncation, they share an inherent limitation: extreme sensitivity to the temperature parameter. Recent logit-space approaches like Top-n\sigma achieve temperature invariance but rely on global statistics that are susceptible to long-tail noise, failing to capture fine-grained confidence structures among top candidates. We propose Min-k Sampling, a novel dynamic truncation strategy that analyzes the local shape of the sorted logit distribution to identify "semantic cliffs": sharp transitions from high-confidence core tokens to uncertain long-tail tokens. By computing a position-weighted relative decay rate, Min-k dynamically determines truncation boundaries at each generation step. We formally prove that Min-k achieves strict temperature invariance and empirically demonstrate its low sensitivity to hyperparameter choices. Experiments on multiple reasoning benchmarks, creative writing tasks, and human evaluation show that Min-k consistently improves text quality, maintaining robust performance even under extreme temperature settings where probability-based methods collapse. We make our code, models, and analysis tools publicly available.

Wenbo Chen, Veena Padmanabhan, Tootiya Giyahchi, Elaine Wong, Leman Akoglu

Hallucination, broadly referring to unfaithful, fabricated, or inconsistent content generated by LLMs, has wide-ranging implications. Therefore, a large body of effort has been devoted to detecting LLM hallucinations, as well as designing benchmark datasets for evaluating these detectors. In this work, we first establish a desiderata of properties for hallucination detection benchmarks (HDBs) to exhibit for effective evaluation. A critical look at existing HDBs through the lens of our desiderata reveals that none of them exhibits all the properties. We identify two largest gaps: (1) RAG-based grounded benchmarks with long context are severely lacking (partly because length impedes human annotation); and (2) Existing benchmarks do not make available realistic label noise for stress-testing detectors although real-world use-cases often grapple with label noise due to human or automated/weak annotation. To close these gaps, we build and open-source a new RAG-based HDB called TRIVIA+ that underwent a rigorous human annotation process. Notably, our benchmark exhibits all desirable properties including (1) TRIVIA+ contains samples with the longest context in the literature; and (2) we design and share three sets of noisy labels with different, sample-dependent noise schemes. Finally, we perform experiments on RAG-based HDBs, including our TRIVIA+, using popular SOTA detectors that reveal new insights: (i) ample room remains for current detectors to reach the performance ceiling on RAG-based HDBs, (ii) the basic LLM-as-a-Judge baseline performs competitively, and (iii) label noise hinders detection performance. We expect that our findings, along with our proposed benchmark, will motivate and foster needed research on hallucination detection for RAG-based tasks.

Riyang Bao, Cheng Yang, Dazhou Yu, Zhexiang Tang, Gengchen Mai, Liang Zhao

Geospatial reasoning is essential for real-world applications such as urban analytics, transportation planning, and disaster response. However, existing LLM-based agents often fail at genuine geospatial computation, relying instead on web search or pattern matching while hallucinating spatial relationships. We present Spatial-Agent, an AI agent grounded in foundational theories of spatial information science. Our approach formalizes geo-analytical question answering as a concept transformation problem, where natural-language questions are parsed into executable workflows represented as GeoFlow Graphs—directed acyclic graphs with nodes corresponding to spatial concepts and edges representing transformations. Drawing on spatial information theory, Spatial-Agent extracts spatial concepts, assigns functional roles with principled ordering constraints, and composes transformation sequences through template-based generation. Extensive experiments on MapEval-API and MapQA benchmarks demonstrate that Spatial-Agent significantly outperforms existing baselines including ReAct and Reflexion, while producing interpretable and executable geospatial workflows.

Victor De Lima, Grace Hui Yang

Most conversational agents (CAs) are designed to satisfy user needs through user-driven interactions. However, many real-world settings, such as academic interviewing, judicial proceedings, and journalistic investigations, involve broader institutional decision-making processes and require agents that can elicit information from users. In this paper, we introduce Information Elicitation Agents (IEAs) in which the agent’s goal is to elicit information from users to support the agent’s institutional or task-oriented objectives. To enable systematic research on this setting, we present YIELD, a 26M-token dataset of 2,281 ethically sourced, human-to-human dialogues. Moreover, we formalize information elicitation as a finite-horizon POMDP and propose novel metrics tailored to IEAs. Pilot experiments on multiple foundation LLMs show that training on YIELD improves their alignment with real elicitation behavior and findings are corroborated by human evaluation. We release YIELD under CC BY 4.0. The dataset, project code, evaluation tools, and fine-tuned model adapters are available at: https://github.com/infosenselab/yield.

Yiqiao Jin, Rachneet Kaur, Zhen Zeng, Sumitra Ganesh, Srijan Kumar

Multi-page visual documents such as manuals, brochures, presentations, and posters convey key information through layout, colors, icons, and cross-slide references. While multimodal large language models (MLLMs) offer opportunities in document understanding, current systems struggle with complex, multi-page visual documents, particularly in fine-grained reasoning over elements and pages. We introduce SlideAgent, a versatile agentic framework for understanding multi-modal, multi-page, and multi-layout documents, especially slide decks. SlideAgent employs specialized agents and decomposes reasoning into three specialized levels–global, page, and element–to construct a structured, query-agnostic representation that captures both overarching themes and detailed visual or textual cues. During inference, SlideAgent selectively activates specialized agents for multi-level reasoning and integrates their outputs into coherent, context-aware answers.Extensive experiments show that SlideAgent significantly improves accuracy over both proprietary (+7.9%) and open-source models (+9.8%).

Michael A. Lepori, Tal Linzen, Ann Yuan, Katja Filippova

Though large language models (LLMs) have enabled great success across a wide variety of tasks, they still appear to fall short of one of the loftier goals of artificial intelligence research: creating an artificial system that can adapt its behavior to radically new contexts upon deployment. One important step towards this goal is to create systems that can induce rich representations of data that are seen in-context, and then flexibly deploy these representations to accomplish goals. Recently, Park et al. 2024 demonstrated that current LLMs are indeed capable of inducing such representation from context (i.e., in-context representation learning). The present study investigates whether LLMs can use these representations to complete simple downstream tasks.We first assess whether open-weights LLMs can use in-context representations for next-token prediction, and then probe models using a novel task, adaptive world modeling. In both tasks, we find evidence that open-weights LLMs struggle to deploy representations of novel semantics that are defined in-context, even if they encode these semantics in their latent representations. Furthermore, we assess closed-source, state-of-the-art reasoning models on the adaptive world modeling task, demonstrating that even the most performant LLMs cannot reliably leverage novel patterns presented in-context. Overall, this work seeks to inspire novel methods for encouraging models to not only encode information presented in-context, but to do so in a manner that supports flexible deployment of this information.

Anneliese Brei, Abhisheik Sharma, Nicholas Sanaie, Lu Wang, Snigdha Chaturvedi

As LLM-generated text is increasingly used, especially in fictional domains, we explore how much LLM-generated stories differ from human-written stories. In this work, we focus on characters. We borrow definitions from narratology to analyze 8 intricate category-pairs of character, such as stylization and wholeness. These category-pairs consider more than just basic characteristics. They assess how characters are portrayed within their stories. After automatically inferring categories of characters within both LLM and human-written stories, we compare and contrast these two sets of stories. We consider the following overarching questions: (1) Do LLMs and human-written stories have similar characters? and (2) Do LLMs generate stories with a variety of characters? Our analysis includes research questions that focus on stories generated by popular LLMs and recently published human-written stories. We describe a number of interesting similarities, differences and key takeaways.

Arda Yüksel, Gabriel Thiem, Susanne Walter, Patrick Felka, Gabriela Alves Werb, Ivan Habernal

Industry classification schemes are integral parts of public and corporate databases as they classify businesses based on economic activity. Due to the size of the company registers, manual annotation is costly, and fine-tuning models with every update in industry classification schemes requires significant data collection. We replicate the manual expert verification by using existing or easily retrievable multimodal resources for industry classification. We present MONETA, the first multimodal industry classification benchmark with text (Website, Wikipedia, Wikidata) and geospatial sources (OpenStreetMap and satellite imagery). Our dataset enlists 1,000 businesses in Europe with 20 economic activity labels according to EU guidelines (NACE). Our training-free baseline reaches 62.10% and 74.10% with open and closed-source Multimodal Large Language Models (MLLM). We observe an increase of up to 22.80% with the combination of multi-turn design, context enrichment, and classification explanations. We will release our dataset and the enhanced guidelines.